Random-effects substitution models for phylogenetics via scalable gradient approximations.

Phylogenetic and discrete-trait evolutionary inference depend heavily on an appropriate characterization of the underlying character substitution process. In this paper, we present random-effects substitution models that extend common continuous-time Markov chain models into a richer class of processes capable of capturing a wider variety of substitution dynamics. As these random-effects substitution models often require many more parameters than their usual counterparts, inference can be both statistically and computationally challenging. Thus, we also propose an efficient approach to compute an approximation to the gradient of the data likelihood with respect to all unknown substitution model parameters. We demonstrate that this approximate gradient enables scaling of sampling-based inference, namely Bayesian inference via Hamiltonian Monte Carlo, under random-effects substitution models across large trees and state-spaces. Applied to a dataset of 583 SARS-CoV-2 sequences, an HKY model with random-effects shows strong signals of nonreversibility in the substitution process, and posterior predictive model checks clearly show that it is a more adequate model than a reversible model. When analyzing the pattern of phylogeographic spread of 1441 influenza A virus (H3N2) sequences between 14 regions, a random-effects phylogeographic substitution model infers that air travel volume adequately predicts almost all dispersal rates. A random-effects state-dependent substitution model reveals no evidence for an effect of arboreality on the swimming mode in the tree frog subfamily Hylinae. Simulations reveal that random-effects substitution models can accommodate both negligible and radical departures from the underlying base substitution model. We show that our gradient-based inference approach is over an order of magnitude more time efficient than conventional approaches.

Random-effects substitution models for phylogenetics via scalable gradient approximations.

Phylogenetic and discrete-trait evolutionary inference depend heavily on an appropriate characterization of the underlying character substitution process. In this paper, we present random-effects substitution models that extend common continuous-time Markov chain models into a richer class of processes capable of capturing a wider variety of substitution dynamics. As these random-effects substitution models often require many more parameters than their usual counterparts, inference can be both statistically and computationally challenging. Thus, we also propose an efficient approach to compute an approximation to the gradient of the data likelihood with respect to all unknown substitution model parameters. We demonstrate that this approximate gradient enables scaling of sampling-based inference, namely Bayesian inference via Hamiltonian Monte Carlo, under random-effects substitution models across large trees and state-spaces. Applied to a dataset of 583 SARS-CoV-2 sequences, an HKY model with random-effects shows strong signals of nonreversibility in the substitution process, and posterior predictive model checks clearly show that it is a more adequate model than a reversible model. When analyzing the pattern of phylogeographic spread of 1441 influenza A virus (H3N2) sequences between 14 regions, a random-effects phylogeographic substitution model infers that air travel volume adequately predicts almost all dispersal rates. A random-effects state-dependent substitution model reveals no evidence for an effect of arboreality on the swimming mode in the tree frog subfamily Hylinae. Simulations reveal that random-effects substitution models can accommodate both negligible and radical departures from the underlying base substitution model. We show that our gradient-based inference approach is over an order of magnitude more time efficient than conventional approaches.

Phylogenetic analysis of adaptation in comparative physiology and biomechanics: overview and a case study of thermal physiology in treefrogs.

Comparative phylogenetic studies of adaptation are uncommon in biomechanics and physiology. Such studies require data collection from many species, a challenge when this is experimentally intensive. Moreover, researchers struggle to employ the most biologically appropriate phylogenetic tools for identifying adaptive evolution. Here, we detail an established but greatly underutilized phylogenetic comparative framework - the Ornstein-Uhlenbeck process - that explicitly models long-term adaptation. We discuss challenges in implementing and interpreting the model, and we outline potential solutions. We demonstrate use of the model through studying the evolution of thermal physiology in treefrogs. Frogs of the family Hylidae have twice colonized the temperate zone from the tropics, and such colonization likely involved a fundamental change in physiology due to colder and more seasonal temperatures. However, which traits changed to allow colonization is unclear. We measured cold tolerance and characterized thermal performance curves in jumping for 12 species of treefrogs distributed from the Neotropics to temperate North America. We then conducted phylogenetic comparative analyses to examine how tolerances and performance curves evolved and to test whether that evolution was adaptive. We found that tolerance to low temperatures increased with the transition to the temperate zone. In contrast, jumping well at colder temperatures was unrelated to biogeography and thus did not adapt during dispersal. Overall, our study shows how comparative phylogenetic methods can be leveraged in biomechanics and physiology to test the evolutionary drivers of variation among species.

Active learning narrows achievement gaps for underrepresented students in undergraduate science, technology, engineering, and math.

We tested the hypothesis that underrepresented students in active-learning classrooms experience narrower achievement gaps than underrepresented students in traditional lecturing classrooms, averaged across all science, technology, engineering, and mathematics (STEM) fields and courses. We conducted a comprehensive search for both published and unpublished studies that compared the performance of underrepresented students to their overrepresented classmates in active-learning and traditional-lecturing treatments. This search resulted in data on student examination scores from 15 studies (9,238 total students) and data on student failure rates from 26 studies (44,606 total students). Bayesian regression analyses showed that on average, active learning reduced achievement gaps in examination scores by 33% and narrowed gaps in passing rates by 45%. The reported proportion of time that students spend on in-class activities was important, as only classes that implemented high-intensity active learning narrowed achievement gaps. Sensitivity analyses showed that the conclusions are robust to sampling bias and other issues. To explain the extensive variation in efficacy observed among studies, we propose the heads-and-hearts hypothesis, which holds that meaningful reductions in achievement gaps only occur when course designs combine deliberate practice with inclusive teaching. Our results support calls to replace traditional lecturing with evidence-based, active-learning course designs across the STEM disciplines and suggest that innovations in instructional strategies can increase equity in higher education.

Systematics of the frogs allocated to Sarcohyla bistincta sensu lato (Cope, 1877), with description of a new species from Western Mexico.

A new species of hylid frog is described from the southwestern edge of the Mexican Plateau from the states of Morelos and Mexico through Michoacán and Jalisco, reaching the Sierra Madre Occidental in Sinaloa and western Durango. The new species is part of the widespread Mexican hylid Sarcohyla bistincta (sensu amplo) complex, comprised of S. bistincta, S. pentheter, S. calthula, and S. ephemera. One subspecies of S. bistincta (labeculata) was proposed for an isolated population in Oaxaca. We restrict the group's nominal species, S. bistincta (sensu stricto), to the Sierra Madre Oriental of Mexico and southward into the Sierra Madre del Sur of Guerrero and Oaxaca. Examination of type material places S. calthula and S. ephemera in the synonymy of S. labeculata (new combination). The species allied to S. bistincta, namely, S. bistincta, S. labeculata, S. pentheter, and the new species described herein, are diagnosed and described following recent suggested taxonomic changes and new available material.

A multilocus phylogeny of the genus Sarcohyla (Anura: Hylidae), and an investigation of species boundaries using statistical species delimitation.

The genus Sarcohyla is composed by 24 species endemic to México. Despite the large number of phylogenetic studies focusing on the family Hylidae, the relationships among the species of Sarcohyla are still poorly known, and the scarce numbers of specimens and tissue samples available for some of the species has hampered an appropriate phylogenetic analysis. We present the most comprehensive molecular phylogenetic study of Sarcohyla to date. We included 17 species of the genus Sarcohyla using data from two mitochondrial (ND1 and 12S) and three nuclear genes (Rag-1, Rhod, and POMc). We performed phylogenetic analyses using Bayesian inference, and the absence of conflicts with strong support between the separate gene trees indicates that incomplete lineage sorting and/or introgressive hybridization are negligible. A coalescent-based species-tree analysis of the four independent loci (three nuclear genes+mtDNA) mostly supports the same species-level relationships as the analysis of the concatenated data. By including new samples from additional species and localities, we find that: (1) the widely distributed species S. bistincta is a complex of at least three species, (2) another undescribed species exists in the group, (3) the species S. ephemera is not valid and it corresponds to a junior synonym of S. calthula. In addition, we conducted marginal likelihood estimation and used Bayes factors to test alternative species delimitation models for S. bistincta, the most widespread nominal species in the group. Our findings support three independent lineages of S. bistincta group, which are paraphyletic with respect to S. pentheter and S. calthula.